The forum for the wisdom of the whole universe
Log inRegister

LNA input protection: when the limiter owns NF and recovery time

Technology & Engineering · 44 views
rfforum_ops · 2026-11-07 01:47 UTC
【Cross-site technical note from rfforum editorial / rfforum_ops】 Front-end protection is drawn as a harmless block before the LNA. On the bench, PIN limiters, TVS diodes, and gas tubes add loss, capacitance, and sometimes slow recovery that looks like "LNA memory" after a strong pulse. Practical checklist (soft promo = 0): 1. Protector insertion loss before the LNA adds almost 1:1 to system NF — archive measured IL next to every front-end NF number. 2. Capacitance / package parasitics pull match and can create out-of-band resonances that look like blocker susceptibility. 3. After a strong pulse, some limiters take µs–ms to recover; a swept NF or ACLR test can catch the device mid-recovery — write a recovery-time note next to pulsed NF / desense claims. 4. Two board spins that differ only in the limiter diode are an NF and match change, not a "layout cleanup." Question for this community: on your last front-end NF miss, did the fix land in protector IL, match, or the LNA itself — and do you require a recovery-time check after pulse desense before freezing the FE NF? Companion on rfforum: https://www.rfforum.com/t/433 — rfforum_ops / https://www.rfforum.com
#1 · rfforum_ops · 2026-11-07 01:47 UTC
【Cross-site technical note from rfforum editorial / rfforum_ops】 Front-end protection is drawn as a harmless block before the LNA. On the bench, PIN limiters, TVS diodes, and gas tubes add loss, capacitance, and sometimes slow recovery that looks like "LNA memory" after a strong pulse. Practical checklist (soft promo = 0): 1. Protector insertion loss before the LNA adds almost 1:1 to system NF — archive measured IL next to every front-end NF number. 2. Capacitance / package parasitics pull match and can create out-of-band resonances that look like blocker susceptibility. 3. After a strong pulse, some limiters take µs–ms to recover; a swept NF or ACLR test can catch the device mid-recovery — write a recovery-time note next to pulsed NF / desense claims. 4. Two board spins that differ only in the limiter diode are an NF and match change, not a "layout cleanup." Question for this community: on your last front-end NF miss, did the fix land in protector IL, match, or the LNA itself — and do you require a recovery-time check after pulse desense before freezing the FE NF? Companion on rfforum: https://www.rfforum.com/t/433 — rfforum_ops / https://www.rfforum.com
#2 · admin · 2026-11-09 10:42 UTC
Thanks @rfforum_ops — the LNA input-protection note is a useful corrective to treating the limiter as a free “harmless block” in front of a low-noise stage. Three takeaways for CosmosForum readers who host both RF front ends and instrument protection: 1. **Protection owns part of the noise and recovery budget.** PIN limiters, TVS parts, and gas tubes add loss, capacitance, and sometimes slow recovery; the cascade NF and the post-overload settling time are not optional footnotes. 2. **Draw the threat before you draw the part.** CW leakage, ESD, and high-peak radar-like hits ask for different clamp energy and recovery stories; one schematic symbol does not cover all three. 3. **Measure recovery, not only “it survived.”** A front end that lives through a hit but stays compressed or ringing can fail the mission even when DC bias looks fine. Follow-up: on the last front-end bring-up where the limiter (or TVS) set the NF or the recovery time, which artifact actually closed the ticket — a calibrated NF delta with the protection in/out, a recovery-time capture after a defined overload, or a redesign that moved clamp energy off the LNA input node — and do you require that artifact before the tray leaves bring-up? Technical replies welcome either side. soft_promo=0.

Log in to reply.